Introduction

Since the mid-1980s, China’s social welfare programs, including pension, health insurance, unemployment insurance, work injury insurance, and the minimum living allowance, have been established and notably expanded. In 2007, more than two hundred million Chinese participated in the social pension and basic health insurance programs. In the same year, the government’s insurance expenditure reached 789 billion RMB and increased by 21.8 % from the previous year (Statistical Communiqué on Civil Affairs Development, 2007). In 2009, the government announced another budget of 850 billion RMB (about 125 billion USD) to set up a health insurance system that would provide all rural and urban residents with affordable health services (2009–2011 Implementation Scheme of Deeper Medical and Health Reform [50]). In retrospect, the Chinese urban social welfare reform that began in the mid-1980s has been ongoing in three distinct dimensions: (1) social insurance: a payroll-tax-based and contributory social insurance system was established that privileges state-owned enterprises (SOEs); (2) privatization: private insurance programs have developed as the self-payment rate for social welfare/services such as health care, education, and housing has skyrocketed; (3) targeting: social security has been established and expanded to urban low-income groups, including laid-off workers, retirees, and unemployed people. Without examining the micro-foundation of each dimension of the reform policies, the trajectory of China’s urban social welfare development seems quite puzzling—it is neither fully liberalized as in Anglo-Saxon countries nor universalized as in Scandinavian countries. One might want to ask questions such as what the politics of Chinese social welfare reform looks like, what are the group preferences and social cleavages underlying China’s social welfare reform, and how these social cleavages and the resulting conflicts influence the reform process and policies. This paper will address these questions, starting from the micro-foundations of China’s urban social welfare reform. It will examine how societal groups aggregate and articulate their welfare preferences and how the Chinese government accommodates them in social welfare reform.

The main argument in this paper is that Chinese societal groups have very different social welfare preferences depending on their employment sectors and status. Based on those diverse social welfare preferences, the politics of Chinese urban social welfare reform since the mid-1980s can be considered as taking place along two lines of division: between labor market insiders and outsidersFootnote 1 and between state sector and non-state sector employees. These two cleavages have shaped the trajectory and policies of Chinese social welfare reform in recent decades. The Chinese authoritarian regime’s political priority of economic growth with social stability has induced the government to accommodate public social welfare preferences in social welfare reform. Establishment of a safety net and expansion of social insurance in urban China can be seen as the government’s main responses to the cleavages derived from public social welfare preferences. To support this argument the paper uses empirical evidence drawn from 80 interviews with enterprise directors, workers, and officials in labor and social welfare bureaus across China between 2009 and 2012.

Literature Review

The welfare state literature is one important component of the study of political economy. The most commonly used explanatory approaches in the welfare state literature are industrialism, institutionalism, and power resources theory. The industrialist explanation maintains that social welfare is a by-product of industrialization and its demographic and social organizational consequences [44, 59]. A contemporary version of the industrialist explanation of the welfare state is globalization theory. Globalization causes domestic vulnerability to external fluctuations and thus incentivizes the state to take over some social responsibilities to mitigate citizens’ increasing risks such as unemployment, aging, disability, and sickness [3, 19, 33]. Contrary to the economic logic emphasized in the industrialist approach, the institutionalist approach stresses the far-reaching and constraining effects of institutions on social welfare change. Scholars holding this view often focus on several variables pertaining to the state, including the policy-making activities and capacities of bureaucrats [28], state centralization [54], state structure [32],Footnote 2 and policy legacy [23, 24]. The institutionalist theory has a strong country and qualitative focus. Huber and Stephens [30] criticize the emphasis on policy legacies and path dependence, which they claim to be a necessary but not sufficient explanation, maintaining that these arguments must be supplemented by power resource theory. With a more agent-oriented perspective, power resource theory, which has been the most influential of the core theories in modern comparative welfare state literature, focuses on the balance of power between labor and capital. In the framework of this theory, the relative power of the left and labor is a strong explanatory factor of welfare state growth [13, 30]. A newly emerging literature that follows the agent-oriented perspective of the power resource theory but more stresses the role of employers and capital in the provision of social insurance has challenged power resource theory. The new literature shows that employers and capital are not uniformly against social insurance programs [14, 40, 52, 53]. Proponents of “varieties of capitalism” contend that private sector actors have their own distinctive social policy interests depending on their underlying production strategies [26].

The aforementioned theories of the welfare state, whether from the structural perspective as in industrialism and institutionalism theories or from the agent-based perspective as in the power resource theory and the employer-based literature, have been geographically confined to the advanced industrial countries, and more particularly to an understanding of the European experiences. Unsurprisingly, the study of social welfare in countries outside the OECD has been a rapidly growing area of research in recent years [1, 5, 25, 47, 48, 55, 56, 60]. Scholars find that economic development and openness alone do not explain much of the variation in social spending among developing countries [48]. The power resource theory is also found to be a less powerful account of social welfare in developing countries because leftist parties and labor organizations are often forbidden or quite weak in many developing countries. While some scholars who study social welfare in the developing world share the institutionalist perspective with their counterparts who study the welfare state in advanced industrialized democracies using formal political institutions such as political parties, legislatures, and bureaucrats as the main explanatory variables of social welfare reform [5, 60], many other scholars emphasize the distributional coalition and its capacities and preferences for social welfare [25, 47, 48]. Rudra [48] points out that the organizing capacity of labor in developing countries is a crucial yet understudied factor in accounting for the welfare regimes in developing countries. She contends that unlike in many of the developed countries, societal groups in developing countries have limited institutional power and their bargaining position is weak, because the sizable population of low-skilled laborers is faced with collective action problems exacerbated by large pools of surplus labor. Likewise, Haggard and Robert Kaufman [25] maintain that institutional rules of the game—the supply side of the political market—are not sufficient to account for the character of social policy without consideration of underlying interests and their organization—the demand side of the political market—and the economic context in which governments operate. They argue that the “critical alignments” that occurred in the mid-twentieth century in different regions of the developing world are important sources of the variety of welfare regimes in those regions [25].

As the literature on social welfare in the developing world surges, studies on Chinese social welfare have proliferated in the past decade as well. Some scholars view Chinese social welfare reforms as one of the social consequences of China’s economic reforms, particularly the SOE reform [2, 4, 21, 34, 42]. Other scholars directly examine the policy-making and implementation of a specific social welfare policy, for example, Frazier’s works on China’s pension policy [15–18] and Duckett’s writings on China’s health and unemployment insurance policies [6–10]. The extant literature on China’s social welfare system has two deficiencies. First, the literature is quite fragmented without much conversation, because of the narrow focus of individual work. Some studies are confined to a certain fragment of the reform periods (e.g., social welfare change in the late 1990s accompanying SOE privatization), while others are confined to a particular social policy area (e.g., pensions, health insurance, or unemployment insurance). Studies of Chinese social policy rarely consider a coherent political economic logic behind the patchwork of Chinese urban social welfare reforms. Secondly, most of the existing studies of China’s social welfare system focus on the supply side of social welfare policies. The supply-driven view dominant in the literature leads scholars often to adopt a “top-down” perspective to comprehend the reform process and policies. In other words, scholars of Chinese social welfare often start the causal chain from the government’s policy initiatives and treat the policy demands and public preferences in the society as given. The aforementioned research on social welfare in countries outside the OECD that emphasizes the importance of the “demand side” stories [25, 48] is a particular source of illumination for Chinese social welfare studies. This paper attempts to fill the gap between the study of Chinese social policy and the comparative welfare state literature by bringing the micro-foundation of Chinese social welfare policy into analysis and exploring the interaction between societal groups and the government during social welfare reforms.

This paper argues that there are four pivotal societal groups in urban China: state sector insiders, state sector outsiders, non-state sector insiders, and non-state sector outsiders. These groups have maintained distinct social welfare preferences in the reform era. State sector insiders are the vested interests who support social welfare benefits being drawn from the contributory and payroll-tax-based social insurance pools; non-state sector insiders prefer more autonomy in social welfare provision and support private insurance programs that better cater to their needs in market competition; state sector outsiders are mostly an older generation of workers who got laid-off during the economic transformation and they call for fair compensation for their losses; non-state sector outsiders, most of whom have migrated from rural to urban areas, need minimum yet universal social welfare to counter risks. The diversity of the societal groups’ preferences for social welfare gives rise to two lines of division of interests in Chinese social welfare reform—between state sector and non-state sector employees and between labor market insiders and outsiders. The division of interests accounts for the politics of Chinese social welfare reforms since the mid-1980s.

The rest of this paper will unfold as follows. “The Trajectory and Policies of Chinese Social Welfare Reform Since the Mid-1980s” will summarize the Chinese social welfare reform policies on three policy dimensions—social insurance, privatization, and targeting. “Societal Groups and Their Preferences for Social Welfare” will examine urban Chinese societal groups’ preferences for social welfare and the two lines of cleavage derived from these diverse preferences during social welfare reform. “The Politics of Social Welfare Reform in Urban China” will present how the Chinese societal groups aggregate and articulate their social welfare preferences; it will also elucidate how and why the authoritarian Chinese government accommodates these demands in social welfare reform. “Concluding Remarks” will summarize the causal logic and conclude the paper.

The Trajectory and Policies of Chinese Social Welfare Reform Since the Mid-1980s

China’s social welfare reform began in the mid-1980s and has been ongoing for more than two decades. The reform is a trial-and-error process with some apparently contradictory policies intertwined. The reform policies can be summarized in terms of three policy dimensions—social insurance, privatization and targeting.

The main component of China’s current social welfare system is social insurance, mainly including pension, health, work injury, and unemployment insurance programs. The last two decades have witnessed the establishment of an employment-based and contributory social insurance system in urban China. In the mid-1980s, experimental measures were adopted by selected cities to create “pooling” of pension funds and to transfer the responsibilities of pension provisions from enterprises to local government. Several regulations issued by the State Council beginning in 1991 formalized the practice of financing pension obligations through a combination of “social pools” and individual retirement accounts [16]. Similarly, after the experiments of “co-payment” and “risk pooling” of health insurance in some cities during the 1980s, in 1994 compulsory social health insurance combined with individual premium contributions began to operate in about sixty Chinese cities. Finally, in 1998, the urban employee basic health insurance system was established. In this system, employers are required to pay 6–8 % of their total payroll into a local health insurance fund (HIF) and dedicate health insurance accounts (HIAs) held in the name of each employee. Employees pay 2 % of their wages into their HIAs. In principal, HIAs pay for an employee’s treatment costing up to 10 % of the local average annual wage, after which the HIF pays. There is a limit on how much the HIF will pay for any single individual, set at four times the average annual wage [8]. As for unemployment insurance, it was initiated in 1986 from the employer-based unemployment insurance established in SOEs that were experiencing restructuring. In 1994, the tripartite contributory unemployment insurance was initiated in the state sector and was extended to the non-state sector in 1999 when the State Council promulgated the Regulations on Urban Unemployment Insurance. According to the 1999 Regulations, employers’ contributions were raised to 2 % of their total pre-tax payroll and individual employees are required to make a premium contribution of 1 % of their wages. By the end of the 1990s, a payroll-tax-based and contributory social insurance system including pension, health insurance, and unemployment insurance had been established in urban China.

Accompanying the establishment of the basic social insurance system in the 1990s, a trend of de facto privatization began as the state gradually retreated from social welfare provision [10, 22]. Since the early 1980s, the government’s overall support for health care and service dramatically decreased.Footnote 3 This is almost equivalent to largely privatizing most Chinese health care facilities, forcing them to rely more on the sale of services in private markets to cover their expenses after allocations from public sources declined. The government has also expressed its intention to develop the market for private insurance. The urban basic health insurance system established since 1998 preserves room for private health insurance so as to pay medical costs above the upper limit of the basic health insurance. In 2004, the government issued regulations on enterprise annuity and encouraged building up enterprise annuity as the second pillar of China’s pension system.

Although Chinese social insurance is principally employment-based, the social welfare reform after the late 1990s began to target labor market outsiders such as SOE early retirees, laid-off workers, and the urban poor, providing them with basic pensions, unemployment benefits, and a subsistence allowance. In 1999, the central government announced the “two guarantees” policy: guaranteeing that laid-off workers receive basic living allowances and that SOE pensioners receive their pensions in full and on time. Accompanying this policy, the responsibility for distributing pensions was transferred from firms to local government agencies such as banks and postal saving institutes. Moreover, the basic health insurance coverage was extended to laid-off workers. Re-employment centers where laid-off workers are supposed to registerFootnote 4 are to make contributions on laid-off workers’ behalf at 60 % of the local average employee wage for the previous year [9]. In 2001, laid-off SOE workers were also entitled to unemployment insurance benefits. Furthermore, the State Council issued the Regulations on the Minimum Living Allowance in 1999. Families whose average income per capita is lower than the local poverty line are entitled to a subsistence allowance. After 1999, unemployment funds could be used to subsidize unemployed persons’ medical treatment [9].

Table 1 summarizes the major social welfare reform policies since the mid-1980s by policy dimension—social insurance, privatization, and targeting. On the whole, China’s social welfare can be considered as expanding mainly along these three dimensions. Consequently, an interesting question arises: why does China’s social welfare reform proceed in such a fragmented way? The next section will address this question by exploring the public’s social welfare preferences and the social cleavages underlying the social welfare reform in urban China.

Table 1 Summary of China’s urban social welfare reform since the 1980s by policy dimension
Full size table

Societal Groups and Their Preferences for Social Welfare

After the market economy reform began in the 1980s, the Chinese workforce began to diversify. Diversification of enterprise ownership and the rise of a free labor market in the mid-1980s were the main forces driving the dramatic change in the Chinese labor force: first, most SOEs went through ownership and corporate restructuring during this period; secondly, non-state sector enterprises, including privately-owned firms, foreign-owned firms, and joint ventures proliferated and grew rapidly; thirdly, the number of rural-to-urban migrant workers continually increased and they have become the main workforce in China’s coastal regions. As a result, the previously uniform industrial workforce became greatly diversified in urban China. Workers’ remunerations and welfare benefits began to vary dramatically across sectors and even within sectors. Meanwhile, a large number of previous SOE workers were laid off or forced into early retirement (nei tui). China’s social welfare reform was mainly responding to the labor dislocation brought about by the market-economic reform in urban areas. To understand the micro-foundation of Chinese urban social welfare policies, this section will examine three crucial aspects of the Chinese urban labor market since the 1980s: first, the formation of different societal groups in urban areas; secondly, diversification of societal groups’ preferences for social welfare; thirdly, the rise of social cleavages as a result of the diverse social welfare preferences of different societal groups.

Employment Status, Sector and Societal Groups

Since the mid-1980s when Chinese economic reform began, the Chinese labor force has been remarkably divided by employment status and employment sector. In this paper, employment status is defined as falling into two categories: labor market insiders and labor market outsiders. Specifically, “labor market insiders” refers to employees with the following characteristics: (1) having a formal employment contract with an enterpriseFootnote 5; (2) receiving remuneration and social insurance benefits (wu xian) including pensions, healthcare, unemployment, work injury, and maternity benefits to which they are entitled by labor laws. Employees who do not fulfill these two conditions are referred as “labor market outsiders.”

Because of the socialist legacy embedded in China’s newly-established market economy, the interests of Chinese labor market insiders are further divided by sectors. The most remarkable division is between the state and the non-state sectors. The state sector includes government institutes and SOEs,Footnote 6 while the non-state sector comprises foreign-owned enterprises (FOEs), privately-owned enterprises (POFs), joint ventures, limited companies, and small private businesses. Chinese labor market outsiders can also be further categorized into two groups according to the sector they used to work for: state sector outsiders and non-state sector outsiders. “State sector outsidersS refers to people who were laid off or who retired early from state sector enterprises, while “non-state sector outsiders” refers to people whose employment was terminated in non-state sector enterprises, who are employed in small private businesses without a formal labor contract, or who are self-employed.Footnote 7 Table 2 summarizes the categorization of societal groups in urban China by the two key factors—employment status and sector.

Table 2 China’s urban societal groups by employment sector and status
Full size table

Different Societal Groups’ Preferences for Social Welfare

The four societal groups (state sector insiders, non-state sector insiders, state sector outsiders, and non-state sector outsiders) in the Chinese urban labor market have very different preferences for social welfare depending on their employment status and sector. This section draws on secondary materials and the author’s 80 interviews with Chinese workers, enterprise directors, and officials in labor and social welfare bureaus between 2009 and 2012 to elucidate the social welfare preferences of the four societal groups.

First, state-sector insiders generally support the payroll-tax-based and contributory social insurances for pension, healthcare, unemployment, and work injury. Actually, they have been the vested interest group in the Chinese social welfare system throughout the reforms. A database complied from 800 state-owned and non-state-owned firms surveyed in 2000 indicates that SOE pensioners in the sample saw a 31.2 % rise in per capita pension benefits between 1994 and 1999, while those retired from other sectors suffered a decline of 15 % ([18], p. 79). Although the economic and social welfare reforms since the mid-1980s have cut off some traditional social welfare provisions for SOE employees such as housing and life-time employment, the risk pools for social insurance have been markedly expanded from individual enterprise to the municipality and even to the provincial level in some cases (such as pensions and health insurance in Beijing, Shanghai, and Ningxia). The expansion of risk pools significantly increases the sustainability of social insurance benefits for SOE employees. Moreover, “socializing” the welfare benefits by means of social insurance substantially reduced SOEs’ operating costs and thus increased their potential for profits, both of which their employees will eventually benefit from. SOE employees, who managed to secure their positions during their enterprises’ restructuring in the late 1990s,Footnote 8 are the winners in China’s urban social welfare reform. Thus, state sector insiders tend to favor the “socialization” of welfare benefits in the form of payroll tax-based and contributory social insurance programs.

On the contrary, non-state sector insiders lack incentives to join the current social insurance programs without extra compensation from the government. China’s non-state sector enterprises concentrate on light and service industries such as food processing, textiles, electronic appliances, and the like. Compared to the large-sized SOEs that are dominant in the monopolistic and strategic industries such as railways, telecommunications, and energy, privately-owned firms in the non-state sector are usually smaller in size and are exposed to more market competition. However, firms in the non-state sector are not all the same. The author’s interviews with Chinese enterprises between 2009 and 2012 find that non-state sector insiders in firms with high profitability such as large foreign-owned firms often receive generous welfare benefits from private insurance programs run by commercial insurance companies (Informants 1 & 2 in the Appendix). However, non-state sector insiders (both workers and managers) in firms with low profitability usually try to avoid participating in any social insurance program (Informants 3, 4, & 5 in the Appendix). Managers (usually also the owners) of small-sized private firms are reluctant to enroll employees in contributory social insurance for fear of increasing labor costs. It is calculated that full insurance coverage (including pension, health insurance, unemployment, work injury, and maternity insurance) will increase labor costs by 3000RMB per worker in Zhejiang Province (Informant 6 in the Appendix).

Workers in the less profitable private firms are usually low-skilled migrants from rural areas with the particular demographic characteristics of being young, healthy, and single. They do not support the payroll-tax-based and contributory social insurances for several reasons. First, because of these workers’ demographic characteristics, they hold less risk-averse preferences for welfare benefits, especially for pension and healthcare. They prefer cash that they can immediately consume to social welfare benefits that they won’t obtain until becoming retired or severely ill. As one factory manager said in an interview,“(migrant workers) are too young to appreciate the value of social insurance” (Informant 7 in the Appendix). Secondly, migrant workers change jobs so frequently that they are usually unqualified for social insurance programs that require premium payments for a certain number of years with one single employer. Thirdly, many of these rural-to-urban migrant workers are unable to settle down in cities due to the discrimination towards the rural population embedded in the Chinese household registration system and the skyrocketing living expenses in urban areas. Migrant workers are uninterested in social insurance programs also because the social insurance funds are pooled at the prefectural and even lower levels (e.g., the county level) so that they “cannot take the entitlements away when changing jobs” (Informant 8 in the Appendix). It is quite common that migrant workers who find new jobs in other cities find that transferring the money in their social insurance accounts from the current region to the new one, if possible, is very time-consuming and costly. Hence, participation in the payroll-tax-based and contributory social insurance without the government’s extra compensation is felt by non-state sector insiders not to be in their best interests and they are not satisfied with the cost-benefit ratio or the geographic restriction designated in the current social insurance programs.

Thirdly, Chinese state-sector outsiders defend the egalitarian social welfare benefits they enjoyed in the “good old days” under the “iron rice bowl.” These workers have experienced a dramatic change of employment status, from being the privileged industrial workers in the socialist era to becoming laid-off workers in the market-economy era, due to the bankruptcies of a large number of SOEs beginning in the mid-1990s. This change not only cut off their income sources but also left them suddenly on their own feet in a society without a well-developed social safety net that provides basic pensions, healthcare, and a minimum living allowance. What is worse, by the Chinese government’s special definition of “laid-off workers” before 2001, these workers were supposed to maintain labor relationsFootnote 9 with their previous employers. As a result, laid-off workers were not eligible for unemployment insurance benefits. Paradoxically, the SOEs that had to lay off or persuade their employees to retire early were also those that could not afford pensions, health insurance, and even severance fees for employees. Consequently, these early retirees and laid-off workers fell into the grey area between the government’s domain and SOEs’ without one agency explicitly in charge of their subsistence allowance and welfare benefits. No doubt state sector outsiders have become China’s most contentious protestors in recent decades [42]. State-sector outsiders who are in their 50s or older today constitute the generation that experienced rounds of political movements and social campaigns since the regime was founded in 1949. They deem social welfare, especially pensions, as the minimum return to their whole life’s services and sacrifices for the new republic’s prosperity. However, the discrepancy between their increasing subsistence needs and their declining political and economic capabilities render them strong-willed but powerless opponents of the current payroll-tax-based and contributory social insurance system.

Finally, non-state sector outsiders prefer universal social welfare benefits. Compared to state sector outsiders who possess decades of work experiences in urban areas, most non-state sector outsiders are relatively young, low-skilled, and geographically mobile (for instance, having migrated from the countryside to various cities). They usually find jobs in small private businesses that need flexible labor but can only offer low wages. These non-state sector outsiders are the “backup” work force for non-state sector enterprises whose production schedules are unstable due to business cycles or external shocks in the international market. The well-being of non-state sector outsiders is contingent on the rise and fall of the world market [29]. With high insecurity of income and job, non-state sector outsiders need basic social welfare provisions such as a minimum living allowance and unemployment insurance benefits (including employment training) that can reduce their risks of falling into poverty in an economic downturn. The author’s interviews with non-state sector enterprises find that non-state sector outsiders share similar social welfare preferences with some non-state sector insiders, especially those who are employed in small private business such as restaurants, construction, and retail. Both groups have difficulties in continually contributing to social insurance payments due to their unstable employment relations and income flows. Hence, they are not interested in the payroll-tax and employment-based social insurance but they support minimum yet universal social welfare and services.

The four societal groups’ social welfare preferences are summarized in Table 3.

Table 3 Social welfare preferences of the four societal groups
Full size table

Cleavages Derived from the Different Social Welfare Preferences

Based on the above four societal groups’ social welfare preferences, two lines of cleavage can be drawn with respect to social welfare politics in urban China: cleavage between labor market insiders and outsiders and cleavage between state sector and non-state sector employees (see Table 4). The magnitude of the cleavages may differ across regions due to local socioeconomic conditions such as sectoral structure and employment situation.

Table 4 Divisions embedded in public preferences for social welfare
Full size table

The insider-outsider cleavage is the most visible one in China’s social welfare politics. Since the Chinese social insurance programs (including pensions and health insurance) are mainly employment-based, the social insurance system inevitably stratifies the labor force by employment status. In addition, China’s private insurance and financial markets are still underdeveloped because of limited openness and various restrictions set by the central government. Chinese labor market outsiders do not have many insurance options outside the social insurance system. As a result, they often end up relying on private savings or family support for living expenses when they are unable to work due to sickness, disability, or aging (Informants 9 &10 in the Appendix). Thus, it is understandable that the insider-outsider divide is likely to escalate into collective actions such as street demonstrations and even violence.

The sectoral divide between state sector and non-state sector enterprises has become more severe as non-state sector enterprises have been proliferating and growing rapidly since the 1990s. State sector enterprises not only differ from non-state sector firms in their respective dominant industries but also in the structure of corporate organization and management. For instance, SOEs still maintain Chinese Communist Party (CCP) branches and in most cases the secretary of the party branch is also the main decision-maker in the enterprise. It is worth emphasizing that most of China’s trade union organizations at the enterprise level exist in the state sector and they are directly “led” by the party branch in enterprises. The party branch in SOEs endeavors to minimize or at least conciliate conflicts between labor and management as well as between enterprises and the government (Informant 11 in the Appendix). The payroll-tax-based and contributory social insurance is well implemented and supported in SOEs not only because it caters to their interests but also because SOEs are less able and willing to resist government decrees. On the contrary, firms in the non-state sector desire more autonomy in management. Within the non-state sector, large-sized and profitable private corporations prefer rewarding their employees by enrolling them in generous private insurance programs such as firm annuities and life insurance, while the small and less profitable private businesses prefer reducing labor costs by minimizing welfare benefits. Neither of these two distinct preferences of non-state sector companies is well met by the current payroll-tax-based and contributory social insurance system. For the private firms with high profitability, the Chinese social insurance programs lack sufficient benefits and flexibility; for the private firms with low profitability, the social insurance programs substantially increase labor costs without desirable benefits. Therefore, unless China’s current social insurance system incorporates the diverse interests of non-state sector enterprises in social welfare policy, such as enhancing benefits and the cross-regional transferability of entitlements, the sectoral divide will continue to be one of the largest obstacles to the expansion of social insurance coverage and sustainability in urban China.

The Politics of Social Welfare Reform in Urban China

Given the diversity of social welfare preferences in urban China, it is essential to understand how the different societal groups have employed the existing political mechanisms or created new ones in the context of Chinese authoritarianism to articulate their policy preferences during social welfare reform. The nature of Chinese social policy-making is basically non-democratic since decision makers are not popularly elected and decision-making usually occurs not through legislation in the congress but through deliberation among relevant party and government agencies. The Chinese policymakers at the central level transmit the policies to local subordinates for implementation. Thus it is also interesting to explore why the policymakers listen to the public and are willing to accommodate societal groups’ preferences in social policy. This section will first examine how the two lines of division in social welfare preferences are articulated by the societal groups and then elucidate how and why the government accommodates public preferences for social welfare in social welfare reform.

The Insider-Outsider Divide and the Establishment of a Social Safety Net in Urban China

As many SOEs were shut down, privatized or converted to shareholding entities in the late 1990s, the Chinese economy underwent a massive labor dislocation. Official employment data shows that SOE employment peaked at about 109.5 million employees in 1995 before falling to 69.2 million at the end of 2002 ([17]:1). The number of workers employed by SOEs in the sectors of manufacturing, mining, and utilities fell from 44.0 million in 1995 to 15.5 million in late 2002, a 65 % decline ([51]: 23). SOEs shed 40.3 million jobs between 1995 and 2002, while non-SOEs, including government and service sector enterprises, increased their workforce by 16.8 million ([17]: 1). Based on a survey of five cities and using international definitions of unemployment, Giles, Park & Zhang [20] found that the unemployment rate among urban residents increased from 6.1 % in January 1996 to 11.1 % in September 2002. Xue & Zhong [63] put the real unemployment rate at 11.6 % in 1999 and 11.5 % in 2002. As for pensions, the rapid decline in SOE workers has turned millions of workers from contributors into pensioners. In Liaoning Province, for example, the number of pensioners rose from 646 thousand in 1989 to 3.15 million in 2003 ([18]: 107).

During this period many workers lost their jobs either by being laid-off or by being forced to retire early when their enterprises were reorganized or privatized. Most of these workers were SOE employees who were already in their 40s or 50s. Despite enjoying the lifetime employment and egalitarian social welfare benefits in the socialist work units for several decades, they suffered low wages and engaged in minimal consumption compared to the younger generation. Pensions or living allowances after retirement are deemed by them as a “symbolic recognition by the state and firms of the former employee’s years of devoted service” ([42]: 350). They are outraged when realizing that after “retirement” their enterprises are actually unable to provide economic compensation as promised in the “iron rice bowl” system. Before 2001, laid-off workers were officially treated differently from “normally” unemployed persons. The artificial distinction rendered hundreds of thousands of laid-off workers in the state sector unqualified for social unemployment insurance benefits. This also created a de jure status cleavage among labor market outsiders, which makes collective action among them harder. Nonetheless, state sectors outsiders, namely SOE retirees and laid-off workers, became the most contentious protestors among labor-market outsiders in urban China in the late 1990s ([15, 31, 42]). Compared to the non-state sector outsiders, SOE laid-off workers and retirees are relatively old, without much concern about family burdens and reputation in the future job market; many have experiences of political participation in Mao’s time; and they are nostalgic because they believe that the “iron rice bowl” system would not have left them standing on their own feet after retirement. In dozens of cities since the early 1990s, some of these pensioners have engaged in protests over unpaid benefits, unfair remuneration in the distribution of severance pay from bankrupt firms, and other causes [42]. Nationwide, the Chinese Ministry of Public Security recorded 8,700 so-called “spontaneous incidents” (such as street demonstrations and riots) in 1993, rising to 11,000, 15,000 and 32,000 in 1995, 1997, and 1999, respectively [45]. In 2003, some 58,000 protests were staged by three million people, including farmers, workers, teachers, and students [39]. Among them, the largest group consisted of 1.66 million laid-off, retired, and active workers, accounting for 46.9 % of the total number of participants that year [46]. Notably in recent years, many of the rural-to-urban migrant workers who mainly comprise the category of non-state sector outsiders also become active participants in collective labor actions such as the “protests against discrimination” [35].

The grievances of labor market outsiders that have led to several episodes of protests pose serious threats to the authoritarian Chinese regime. The period from the late 1990s to the early 2000s is particularly politically sensitive for the regime because the nation not only had to deal with the Asian economic crisis but also went through mass and painful SOE privatization domestically. Thus, labor issues arising in economic crises were the Achilles’ heel of China’s social stability (Informant 12 in the Appendix). The CCP makes social stability a priority target with veto power in local cadre evaluation.Footnote 10 Local officials who fail in securing social order in their jurisdictions are likely to fail in promotion no matter how well they perform in other policy areas [11, 12, 58]. For this reason, local officials who by design would become the immediate victims of labor unrest are very cautious and pay close attention to social welfare policies. Since the late 1990s, the Chinese government has adopted various measures targeting labor-market outsiders, particularly the contentious SOE retirees and unemployed workers, to prevent social unrest.

The government’s efforts in responding to labor market outsiders’ social welfare demands can be seen in two specific measures: expanding unemployment insurance coverage and establishing the social safety net. As mentioned above, laid-off SOE workers were ineligible for social unemployment insurance benefits because they were still considered by the government as keeping labor relations with their previous work units. In 2001, the re-employment service centers in SOEs and industrial cities that were established to manage laid-off workers affairs and maintain labor relations with them were eliminated. Since then, laid-off workers have been placed in the social unemployment insurance pools. This is a milestone in eliminating the discrimination against laid-off workers in unemployment insurance benefits. Besides, the unemployment insurance coverage was extended beyond the state sector and eventually made mandatory for all urban employers and their employees according to the National Unemployment Insurance Regulations in 1999. At present, migrant workers, after a period of employment, are also entitled to the same re-employment services on the same terms as those provided to their urban counterparts ([9]: 7). The dramatic extension of urban unemployment insurance coverage (at least in principle) is the state’s notable response to the status divide between laid-off workers in the state sector and ordinarily unemployed persons in the non-state sector. It also mitigates the labor market insider-outsider cleavage because around 20 % of the unemployment insurance funds are used to pay for re-employment training and job introduction services for labor market outsiders. In 1999, the same year the unemployment insurance coverage was enlarged, the Regulations on the Minimum Living Allowance for Urban Residents were promulgated. This policy aims to provide a safety net with a subsistence allowance to labor-market outsiders whose income falls beneath the local minimum subsistence level. Starting in Shanghai and Tianjin in the 1990s, this safety net for the poorest began to be extended nationwide.

The Sectoral Divide and the Expansion of Social Insurance Coverage in Urban China

The sectoral divide among labor market insiders is manifest in the expansion of social insurance coverage. SOEs have been enjoying a lot of advantages in the Chinese political economy. First, large-sized SOEs monopolize the production and markets of strategic industries with very high profitability, such as energy, telecommunications, and transportation. Secondly, many large sized SOEs are politically powerful because their management possesses high bureaucratic ranking. Many of these personnel possess vice-ministry or even ministry ranks that are much higher than those of local municipal officials. These bureaucratic rankings, though a legacy of the command economy in socialist times, constantly insulate these companies from the supervision of local authorities. Thirdly, SOEs also keep close ties with government agencies and bureaucrats. They usually seek protection from the National Development and Reform Commission (NDRC) at various levels, and the NDRC’s officials often represent the interests of SOEs in deliberations and disputes over share issuance, corporate reorganizations, bankruptcy proceedings, and taxation and social insurance collections.

The Chinese government’s policy-making has been characterized as “fragmented authoritarianism” by many scholars [36, 49, 57] for two reasons. First, the policy-making process is opaque in China compared to democratic countries. Most policies are made by the government in the form of State Council decrees, ministry regulations, or internal notices. Secondly, political actors within the central government involved in policy-making are limited and fragmented. Generally speaking, decision-makers include the core elites of the top leadership who usually publicly announce or support new policy initiatives, some general coordinating commissions such as the NDRC (usually through its subordinate ad hoc commissions or research institutes), and, depending on policy areas, some ministries of the State Council with their research centers. These individuals and agencies have their own interests and organizational missions that are implicitly related to the different “constituents” whose interests they strive to defend in the policy-making process [36, 49, 57]. For example, the current Ministry of Human Resources and Social Security used to be the Ministry of Labor and Personnel before the economic reform and became the Ministry of Labor and Social Security during the market economy reform. It has been in charge of labor and social insurance affairs for the state sector for a long time. This Ministry tends to defend the interests of state sector enterprises, most of which are SOEs [61]. Due to their long-lasting and close relations with the government, SOEs’ interests are very likely to be over-represented in government policy-making. In fact, protecting SOEs from increased market competition by lightening their burdens of social welfare provisions was the original intention and has always been a major goal of China’s social welfare reform [38].

Since the social insurance system was established in the 1990s to socialize risks and maintain welfare benefits for the state sector, the Chinese government has kept increasing the benefit level of social insurance. For example, even though the number of retirees nation-wide grew from 34.4 million to 37.7 million (by around 10 %), the reported annual per capita pension distribution rose by more than 15 % between 1998 and 2000 ([16]: 58). The average reimbursement rate of health insurance for urban employees was set by the central government at above 85 % in 2011 (Informant 13 in the Appendix). Frazier attributes the Chinese government’s incentives for social insurance expansion to that “(pension) funds have become an important resource for politicians to deliver selective benefits to specific constituencies.” ([18]: 20) The question is where the money will come from to support the dramatically increased expenditures on social insurance. It is found that there are three main sources of funds. The first source is the central government itself, which keeps injecting fiscal resources into the social insurance funds (Informants 14 & 15 in the Appendix). The second as well as the most controversial source is mandatory individual accounts in Chinese social insurance programs (mainly pension and health insurance). China’s pension and health insurance are designed to be financed by both social insurance funds and mandatory individual accounts. It is said that local governments have already emptied the “notional” individual accumulated accounts of current workers to cover the pension and health care distributions for current claimants ([18]: 13). The third source for funding the increasing benefits of social insurance is significant contributions from the non-state sector. In order to increase insurance contributions, the government eagerly and strategically extends social insurance coverage to non-state sector enterprises. In the author’s interviews, informants reveal various local “tricks” of expanding social insurance coverage (kuo mian) to the non-state sector:

Private manufacturing companies are very interested in purchasing social work injury insurance because they find it quite costly to appease injured workers after accidents happen. Accordingly, local labor and social security officials work out a policy combining work injury insurance payment with health insurance and pension payments. Thus, the companies have to complete payments for social health insurance and pensions in order to be qualified for work injury insurance benefits. This strategy significantly increases the enrollment rates of pension and health insurance. (Informant 16 in the Appendix)

The local rule for non-state sector firms participating in social insurance is that at least 50 % of their employees should be enrolled in social insurance programs. As long as enrollment rate is above 50 %, the firm is generally safe in labor supervision. (Informant 17 in the Appendix)

Given the striking divide between the state sector and the non-state sector in China’s political economy, political exchange is indispensable in promoting social welfare reform. Some scholars state that China’s social insurance, especially pension and health insurance, is in a deficit by any measure [8, 18]. Under the constraints of fiscal austerity, social welfare reform is not a win-win game but has to be a zero-sum one [27]. However, the Chinese government plays strategically, changing the nature of the game. The government’s strategy is to incorporate non-state sector enterprises into the social insurance pools as potential contributors; in exchange, the government has made incremental reforms in the social insurance system to accommodate some of the non-state sector’s social welfare preferences. The reform policies along this line can be seen twofold. First, the government has gradually capitalized a portion of the insurance funds and encouraged developing private insurance programs as supplements to basic social insurance. This meets the social welfare preferences of profitable and large-sized enterprises in the non-state sector. In 2000, the central government established the National Social Security Fund (NSSF). As of 2005, it had about 180 billion RMB, and the assets were invested by ten fund management companies that acted as trustees on the board of the National Council of the Social Security Fund ([18]: 64). Moreover, it was reported that in 2002 there were 29 life insurance companies and 8 property insurance companies providing over 300 health insurance products in China (Informant 18 in the Appendix). Most of them are for serious diseases and medical expense insurance, which are supplementary to the basic social health insurance. In 2004, the Chinese government issued regulations to provide a legal framework for voluntary corporate pension plans known as enterprise annuities, the so-called “second pillar of China’s pension system” [62]. In 2009, Fujian Province in southeast China announced that by May 2009, 4,695 enterprises in its jurisdiction had established enterprise annuities covering 250.1 thousand employees [64]. In addition to enterprise annuities, a handful of commercial insurance companies are allowed to offer pension insurance contracts, which account for approximately 47 % of enterprises’ supplementary pension plans in China [62].

The second set of means by which the government adapts social welfare policy for the non-state sector is making special arrangements such as lower contribution rates, shorter periods of premium payment, and ad hoc province-to-province agreement on social insurance transfers to include the non-state sector enterprises whose employees are mostly migrant and temporary workers into social insurance pools. For example, Shanghai labor authorities have attained broad pension coverage rates by essentially allowing different sectors of the labor market to contribute at different rates ([18]: 116). In 2010, 45.83 million migrant workers were enrolled in the social health insurance program for urban employees at lower contribution rates (Informant 19 in the Appendix). The incorporation of millions of young and healthy migrant workers into the urban social insurance pool significantly improves social insurance’s risk structure and financial sustainability. Meanwhile, many non-state sector firms still seek to avoid social insurance contributions because under China’s fragmented and authoritarian policy-making structure, the social welfare provision will continue to be skewed to the state sector. The current social insurance reimbursement continues to reward long-term and stable employment, which is often not the case for non-state sector firms. Fairness of social insurance benefits across sectors and employment status is still a serious problem in the Chinese social insurance system.

The three dimensions of Chinese social welfare reform policies since the mid-1980s that were summarized in “The Trajectory and Policies of Chinese Social Welfare Reform Since the Mid-1980s” quite significantly reflect the societal groups’ different social welfare preferences (see Table 5).

Table 5 Social welfare reform policies accommodating different societal groups’ social welfare preferences
Full size table

The Chinese Government’s Responsiveness to Public Social Welfare Preferences

So far, this paper has demonstrated how Chinese social welfare reform policies in the past few decades correspond to the underlying social cleavages. One might raise the question of why Chinese policymakers are well aware of public reaction and consider it as a potential constraint on policy choice, given that without democracy they are relatively well insulated from organized social forces that could apply the sorts of interest group pressures that exist in pluralist political systems. This paper finds that the three dimensions of Chinese social welfare reform policies—social insurance, privatization, and targeting—do reflect and respond to some of the societal groups’ social welfare preferences that have been articulated in the aforementioned formal and informal channels during the reform. Some other scholars’ studies of Chinese social welfare also concur with this finding. The creation of social insurance-based welfare policies such as pensions shows that the Chinese government can be responsive to social forces even without formal institutions of representative government [18]. This can be attributed to the government’s desire for social stability.

Although some progress has been made with regard to political institutionalization and accountability in China [42, 65], crucial policy-making is not open and is often carried out among the elite, mainly within the CCP central committee. The CCP’s legitimacy deriving from the revolution has gradually died away and come to rest upon a fragile base of bringing improvements to the well-being of the people. Despite the fact that the CCP has succeeded in economic development through decades of incremental market economy reforms, it still lives under the shadow of the future—vulnerable to existential challenges that mature democratic systems do not face. Therefore, maintaining economic growth with social stability has been and will continue to be the CCP central leadership’s political priority. Accommodating societal groups’ social welfare preferences is not adverse but necessary to achieving this goal.

Moreover, the Chinese decentralized political system facilitates the accommodation. The central government actively makes use of various channels including government-sponsored conferences, seminars, surveys, and investigations to collect and understand public preferences for social welfare. More importantly, local officials and cadres who know public preferences better than the central government does are given much discretion in reform policy experiments and implementation within a certain framework:

As the central government determines to expand social insurance coverage, local officials, who desire to meet the target set by the upper level, have to deal with those private and small businesses that do not want to join social insurance. Through daily contact and negotiation again and again, local cadres know about those groups’ problems and concerns about social insurance. They try to work out feasible policies to induce more societal groups to enroll in social insurance. (Informant 20 in the Appendix)

Furthermore, the Chinese government’s desire to incorporate different societal groups into social insurance meets the inherent expansive logic of social insurance. As long as the social insurance does not fully cover the whole population, stakeholders of the social insurance have an interest in enrolling more people into it to share contribution burdens and enhance risk pools. Besides, the government’s remarkably increasing fiscal revenues in recent years enable social expenditure to expand as well.Footnote 11

Concluding Remarks

Chinese social welfare reform since the mid-1980s has been characterized as incremental and fragmented. This paper has attempted to explain the politics underlying the multi-dimensional Chinese social welfare reform policies. Chinese social welfare reform has taken place mainly in three dimensions—social insurance, privatization, and targeting. Without understanding the public social welfare preferences, the patchwork of Chinese social welfare reform policies looks puzzling and contradictory. This paper argues that there are four pivotal societal groups (state sector insiders, state sector outsiders, non-state sector insiders, and non-state sector outsiders) with very distinct social welfare preferences in urban China. The diversity of the societal groups’ preferences for social welfare has given rise to two lines of division of interests in urban China—between state sector and non-state sector employees and between labor market insiders and outsiders. The labor market outsiders who suffer from the insider-outsider conflict articulate their social welfare preferences through collective actions and are eventually appeased by the newly established social safety net including the minimum living allowance, unemployment and pension benefits. The interests of non-state sector enterprises, who have long been disadvantaged in the “Chinese fragmented authoritarianism,” are partially accommodated by recent social welfare reforms in terms of privatization combined with expansion of social insurance pools.

One might be left wondering whether the argument of this paper applies to rural China as well. During the past decades, China had been upholding a rural–urban divide, with exclusive and privileged social welfare provision to urban workers. As urban social welfare provision has recently expanded to non-state sector and labor market outsiders, more and more rural people, especially the peasants who have lost land during urbanization and the rural young generation who migrate to cities for jobs, are incorporated into the urban social insurance system. By 2012, 41 prefectural cities and 5 provinces have abolished the rural–urban divide in health insurance enrollment (Informant 21 in the Appendix). Many other cities are considering merging the rural cooperative medical insurance (xin nong cun he zuo yi liao) into the urban resident health insurance in the near future. Like the social welfare reforms in urban areas during the past few decades, the Chinese government will have to handle the tensions among different societal groups (e.g., people with and without land) in rural areas to develop a national social welfare system that is compatible with rapid urbanization and socioeconomic development. Ultimately, the process of Chinese social welfare reforms illuminates ways in which the Chinese political and economic systems can progress in the twenty-first century.